1. Field of the Invention
The present invention relates to a motion information processing method and a motion information processing apparatus for use in a video processing apparatus that handles video information of an image-pickup device and a display device. More specifically, the present invention relates to a method and an apparatus for processing image motion information in the conversion of interlace-scanned video information to progressive-scanned video information.
2. Description of the Related Art
A technique called interlace-progressive (IP) conversion is known. With that technique, interlace-scanned video information is converted to progressive-scanned video information when video is presented in progressive scanning rather than in one-line interlace scanning.
One type of IP conversion techniques, known as motion adaptive IP, switches between an inter-field interpolation and an intra-field interpolation in response to the motion of an interpolation pixel to be interpolated. In the inter-field interpolation, a value of a video signal (an interpolation value) of the interpolation pixel is determined from a (reference) video signal of a pixel of an image in an adjacent field, at the same location as an interpolation pixel. In the intra-field interpolation, an interpolation value is determined from video signals of two reference pixels in two locations at adjacent upper and lower lines within the same field as the field of the interpolation pixel. In the motion adaptive IP conversion, the interpolation value may be calculated by modifying the weighting of an inter-field interpolation value and an intra-field interpolation value in accordance with the amount of motion of the interpolation pixel.
The known IP conversion uses a difference value of the video signals of pixels at the same location between consecutive odd fields (or consecutive even fields) to detect motion of the pixels. If an motion occurs between frames, as in an even field between two odd fields, a moving picture may be mistaken for a still picture. An erroneous interpolation value thus results.
Japanese Patent Laid-Open No. 2000-175159 discloses a technique that references not only motion information of a target field but also motion information in a field prior to the target field.
Japanese Patent Publication No. 8-32025 discloses a technique that adds, to motion information of a target field, a value that is obtained by multiplying an attenuation coefficient to motion information detected in a field prior to the target field. In accordance with that technique, the effect of the motion information detected as a moving picture persists across a plurality of fields.
There are some cases where the technique disclosed in Japanese Patent Laid-Open No. 2000-175159, in other words, simply referencing the motion information of the prior field is unable to overcome the drawback of the erroneous interpolation value.
In such cases, according to the study of the inventors of this invention, the storing of the motion information in the past field is effective, but the motion information of all pixels can be processed as being a moving picture if the past motion information is merely stored and used. In this respect, the technique disclosed in Japanese Patent Publication No. 8-32025 is found to be effective.
When input video is switched to a still picture, the motion information that is detected as a moving picture in a past field still has an effect even though the attenuation coefficient is multiplied. When an interpolation value must be calculated as a still picture in a given field, an interpolation component may be calculated as a moving picture instead. Replaying an precise image remains difficult until the interpolation component as the moving picture is substantially nullified.